Is the continuity of externalizing psychopathology the same in adolescents and middle-aged adults? A test of the externalizing spectrum's developmental coherence.

Abstract

Externalizing psychopathology (EXT) is a framework for understanding diagnostic comorbidity and etiology of antisocial and substance-use behaviors. EXT indicates continuity in adulthood but the structure of adolescent EXT is less clear. This report examines whether adolescent EXT is trait-like, as has been found with adults, or categorical. We use tests of measurement invariance to determine how diagnostic indicators of EXT differ in adolescents compared to adults. The EXT measures employed were DSM-IIIR diagnoses of adult antisocial behavior, conduct disorder, and alcohol, marijuana, and drug dependence. Latent trait, latent class, and hybrid models were fit to two separate data sets: 2,769 seventeen-year-old adolescents and 2,619 adults from the Minnesota Twin Family Study. The best model in both samples was a single-trait LT model. Parameters from the adolescent and adult models were equivalent for all disorders except alcohol dependence. It appears that EXT in adolescence can be accurately represented by a single-trait model, and the measurement properties of EXT are similar during these time periods with the exception of alcohol dependence.